CN111409648B - A driving behavior analysis method and device - Google Patents

Abstract

The invention provides a driving behavior analysis method and a driving behavior analysis device, wherein the method comprises the following steps: acquiring driving behavior data of a driver aiming at a preset simulated traffic scene; processing the driving behavior data to obtain key index data of a simulated traffic scene; constructing a driver simulation model according to the key index data; and (4) performing anthropomorphic processing on the automatic driving algorithm to be analyzed by utilizing the driver simulation model. Based on the method disclosed by the invention, the automatic driving algorithm can be learned and the human driving behavior can be simulated as much as possible, thereby being beneficial to improving the safety of the automatic driving automobile and being popularized comprehensively.

Description

A driving behavior analysis method and device

technical field

The present invention relates to the technical field of autonomous vehicles, and more particularly, to a driving behavior analysis method and device.

Background technique

With the rapid development of autonomous driving technology, the proportion of cars with different degrees of automation on public roads is increasing.

Since vehicle automation is a slow and gradual process, there will be a mixture of human drivers and autonomous vehicles in the traffic environment for a long time. Human drivers have developed subtle driving behaviors and can understand each other in interactions. However, due to the limitation of the working mode, this way of observation, understanding and communication of human beings cannot be directly accepted by the computer. Therefore, how to make autonomous vehicles safely adapt to the mixed traffic environment of human drivers and autonomous vehicles is an urgent problem to be solved at this stage.

SUMMARY OF THE INVENTION

In view of this, in order to solve the above problems, the present invention provides a driving behavior analysis method and device. The technical solution is as follows:

A driving behavior analysis method, including:

Obtain the driving behavior data of the driver for the preset simulated traffic scene;

processing the driving behavior data to obtain key indicator data of the simulated traffic scene;

Build a driver simulation model according to the key indicator data;

Use the driver simulation model to anthropomorphize the automatic driving algorithm to be analyzed; wherein,

The obtaining of the driving behavior data of the driver for the preset simulated traffic scene includes:

Determine the simulated traffic scene specified by the driver based on the preset environment interactive interface;

Activate pre-set cockpit simulators;

The driving behavior data generated by the cockpit device is collected.

Preferably, the construction of a driver simulation model according to the key indicator data includes:

Determine car following behavior data based on the key indicator data, and process the car following behavior data to obtain a car following behavior model; and/or

The lane-changing behavior data is determined based on the key indicator data, and the lane-changing behavior data is processed to obtain a lane-changing behavior model.

Preferably, the method further includes:

A traffic environment simulation model is constructed based on the driver simulation model.

Preferably, the method further includes:

The automated driving algorithm after anthropomorphic processing is tested by using the traffic environment simulation model.

A driving behavior analysis device, comprising:

an acquisition module, used to acquire the driving behavior data of the driver for the preset simulated traffic scene;

a first processing module, configured to process the driving behavior data to obtain key indicator data of the simulated traffic scene;

a building module for constructing a driver simulation model according to the key indicator data;

The second processing module is used for anthropomorphizing the automatic driving algorithm to be analyzed by using the driver simulation model; wherein,

The acquisition module is specifically configured to: determine the simulated traffic scene designated by the driver based on the preset environment interactive interface; activate the preset cockpit simulation device; and collect the driving behavior data generated by the cockpit device.

Preferably, the building module is specifically used for:

Determine car following behavior data based on the key indicator data, and process the car following behavior data to obtain a car following behavior model; and/or determine lane change behavior data based on the key indicator data, and process the lane change behavior data to obtain Lane changing behavior model.

Preferably, the building block is also used for:

A traffic environment simulation model is constructed based on the driver simulation model.

Preferably, the second processing module is also used for:

The automated driving algorithm after anthropomorphic processing is tested by using the traffic environment simulation model.

Compared with the prior art, the beneficial effects realized by the present invention are:

The invention provides a driving behavior analysis method and device. The method can construct a driver simulation model based on the driving behavior data of the driver in the simulated traffic scene, and then use the driver simulation model to perform anthropomorphic processing on the automatic driving algorithm to be analyzed. . Based on the method disclosed in the present invention, it is possible to realize the automatic driving algorithm learning and imitating human driving behavior to the greatest extent possible, thereby helping to improve the safety and comprehensive promotion of the automatic driving vehicle.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only It is an embodiment of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to the provided drawings without creative work.

1 is a method flowchart of a driving behavior analysis method provided by an embodiment of the present invention;

Figure 2 is an example of a simulated traffic scene;

3 is a flowchart of another method of a driving behavior analysis method provided by an embodiment of the present invention;

FIG. 4 is a schematic structural diagram of a driving behavior analysis device provided by an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The present invention provides a driving behavior analysis method. The method flowchart of the method is shown in FIG. 1 , and includes the following steps:

S10: Acquire driving behavior data of the driver for a preset simulated traffic scene.

In the process of executing step S10, the driving behavior data of the driver for the designated simulated traffic scene may be acquired based on the preset environment interactive interface and the preset cockpit simulation device. Specifically, the simulated traffic scene designated by the driver on the preset environment interactive interface can be determined, the cockpit simulation device can be further activated, and the driving behavior data generated by the cockpit device can be collected. The simulated traffic scene is a pre-created virtual traffic environment, including virtual road conditions, virtual motion states of traffic participants, and virtual weather conditions.

All vehicles and road environments in the simulated traffic scene are virtual mathematical models, and the vehicles can be divided into two categories: the first category is the vehicle controlled by the human driver through the cockpit simulation device (Figure 2 shows the simulated traffic scene In the example, the "host car" that brakes and avoids, that is, the right-hand vehicle); the second category is the vehicle controlled by a non-human driver (as shown in Figure 2, the "front car" that brakes and decelerates in the simulated traffic scene example is shown in Figure 2. ”, i.e. the vehicle on the left). in,

For the first type of vehicle, the driver sends real control signals to the vehicle model through the cockpit simulation device, such as actual manipulation behaviors such as accelerator, braking, steering, etc., so as to control the vehicle's following, acceleration, braking and other behaviors accordingly. . In addition to the basic vehicle control mechanisms (such as accelerator pedal, brake pedal, steering wheel, shift lever, paddles related to various driving functions, etc.) in the real cockpit, the cockpit simulation device is also equipped with The data acquisition sensor at the mechanism, for example, the accelerator pedal position sensor is installed at the accelerator pedal, which can sense the accelerator opening degree manipulated by the driver in real time.

For the second type of vehicle, it can simulate the driving behavior of human drivers, including the perception of the environment, such as identifying other vehicles within a certain range, identifying other traffic participants such as pedestrians, and judging their motion status, such as speed, direction , relative distance, etc., so as to issue control commands to reasonably control speed, acceleration, etc., so as to realize virtual control and simulate a near-real traffic environment.

Taking the "following car" behavior in driving behavior as an example, the driving behavior data is shown in Table 1. The data in the table are all time series during the driving process, the sampling frequency is f (Hz), and the corresponding sampling period is Δt =1/f(s).

Table 1

It should be noted that Table 1 only illustrates the driving behavior data that can be collected in this embodiment. In the driving behaviors that are different from those of following the car, more other types of driving behavior data can be obtained. At the same time, the data generated by the object , and is not limited to the "front vehicle" and "host vehicle" listed in Table 2. In practical applications, multiple "environmental vehicles" and multiple "host vehicles" may also be involved. In this case, the data types of the collected driving behavior data will increase accordingly.

It should also be noted that the collected driving behavior data may be stored in a multi-dimensional logical structure such as a time series, a driver number sequence, and a scene sequence.

S20, processing the driving behavior data to obtain key indicator data of the simulated traffic scene.

In the process of executing step S20, different key indicators may be set for different driving behaviors in different simulated traffic scenarios.

Continuing to take the “car following” behavior in driving behavior as an example, the key indicators shown in Table 2 are calculated based on the driving behavior data. ), the behavioral characteristic parameters of the human driver who controls the host vehicle.

It should be noted that the classification of driving behaviors covered by simulated traffic scenarios includes various behaviors that may occur during actual driving, such as following a car, changing lanes, overtaking, turning, avoiding pedestrians, etc.

Table 2

For the key indicators defined in Table 2, the specific calculation method is shown in Table 3. The coincidence and operators involved in Table 3 are explained as follows: time refers to time, t refers to t time step, and Δt refers to time step (sampling period). The operator | refers to "when the conditions are met", max() refers to the maximum value, min() refers to the minimum value, and || refers to the absolute value.

table 3

S30, constructing a driver simulation model according to the key indicator data.

In the process of executing step S30, the driver simulation model may be directed to vehicle following behavior and lane changing behavior. The other behaviors are, in essence, a combination of the above two decomposition behaviors. For example, overtaking is to follow the car in this lane first, then change lanes, and then follow the car in front of the target lane; another example, entering the ramp and exiting the high-speed, is essentially changing lanes and then following the car.

For the following model corresponding to the following behavior, the acceleration control method can be used to process the following behavior data in the key index data. The acceleration control method is as follows:

The definitions and units of each parameter are shown in Table 4 below:

symbol definition unit x Location m V speed m/s V₀ desired speed m/s T Expected safe headway s a maximum acceleration m/s² b comfortable acceleration m/s² δ acceleration index \ s₀ Parking distance m

Table 4

The parameters in Table 4 can be calibrated by driving behavior data to obtain a model that is as close to the real human driving behavior as possible.

For the lane-changing model corresponding to the lane-changing behavior, the "condition-motivation-execution" control method can be used to process the lane-changing behavior data in the key index data.

According to the actual driving behavior characteristics of human drivers on the road, the lane changing habits of human drivers are divided into four categories: target speed driving, speed keeping + overtaking driving, lane keeping driving, and maximum speed driving.

Target speed driving, this type of driver, will choose a certain desired speed and maintain this speed as the main driving goal. If the driving environment causes them to actually drive slower than expected, they will take every opportunity to overtake as much as possible and will not take the right-hand rule into account when making lane change decisions. If there are no conditions that would cause them to slow down, they will not actively change lanes.

Speed maintenance + overtaking driving, similar to the previous type, this type of driver will also set a desired speed, but the difference is that they will implement the principle of staying on the right, only use the left overtaking lane when overtaking, and will only use the left overtaking lane when overtaking is completed. will return to the right lane.

Lane Keeping Driving, this type of driver tends to stay in the current lane, they will try to adapt to the speed of the car ahead in the current lane as much as possible. Since their main goal is to "keep the lane", they differ from the previous two types of drivers mainly in that they do not have a desired speed, or that they can tolerate a wider range of speeds than the previous two types of drivers.

Maximum speed driving, this type is similar to the third type in that this type of driver does not set a desired speed, but drives as much as possible at the maximum driving speed that can be satisfied by the current driving conditions. Therefore, the lane-changing characteristic of this type of driver is to overtake in order to obtain the maximum speed possible under any conditions that satisfy the overtaking conditions or allow them to gain a speed advantage.

For the above four types of drivers, during the whole process of changing lanes, first of all, the driver will constantly observe the driving environment he is currently in, compare and evaluate it with his own expectations, and come to different degrees of "lane changing". "Motivation", "lane change motivation" is divided into three levels of different degrees, the first level is "not required", when the driver evaluates the result as "not required", keep the current lane, the second level is "generally required" ”, when the evaluation result falls at this level, the driver will start to look for a lane-changing opportunity, if there is a safe enough and suitable lane-changing opportunity at this time, execute the lane-changing, otherwise, keep the current lane, and re-evaluate the “lane-changing opportunity” The third level is "strong need". In this level, the driver has an urgent need to change lanes. The driver will look for an opportunity to change lanes. If it is safe, change lanes. Then wait for an opportunity or create an opportunity (such as slowing down or accelerating) until a lane change condition is available, and then execute a lane change.

For the lane-changing model, another dimension that affects the motivation of lane-changing needs to be considered—the road condition factor. The conditions faced by drivers that affect their "lane change motivation" are divided into seven categories, as follows:

Ready to turn: Needs to turn after a certain distance and is not in the correct turn lane now.

Lane Termination: The lane you are currently in will terminate within a certain distance.

Borrowed lane: temporarily in the wrong lane for temporary reasons, for example, temporarily in the right lane after turning right, but the next intersection needs to go straight.

Unexpected obstacle: An unexpected obstacle is encountered in front of one's own lane.

Speed advantage: Changing lanes allows you to increase your speed.

Queue advantage: When you are stuck in traffic or waiting for a signal, changing lanes can help you get ahead in the queue.

Advantages of getting out of congestion: When there is congestion ahead, changing lanes can make him get out of the congested road faster. For example, when the driver notices that the congestion is caused by vehicles merging in on the right side ahead, he tends to change lanes in advance Go to the far left lane, as this will help reduce his time in congested areas.

Among them, in the first four cases, the corresponding lane changing needs are "strong needs", and the last three cases correspond to "general needs".

S40, using the driver simulation model to perform anthropomorphic processing on the automatic driving algorithm to be analyzed.

In the process of executing step S40, the process of anthropomorphic processing is to first conduct statistics on the driving behavior of human drivers for specific control parameters, such as the distance following the vehicle, and the human drivers are used to Taking the following distance of f(v), then the following distance of f(v) is used as a reference to adjust the "following distance" defined in the intelligent vehicle control algorithm.

It is worth noting that the "anthropomorphic processing" discussed above is only for "non-emergency scenarios", or, in other words, limited to scenarios that excellent human drivers can deal with. That is to say, this part is only used to evaluate The "comfort" and "efficiency" of smart cars under the premise of ensuring "safety", while another part of the scene is beyond the ability of human drivers to deal with, or the degree of urgency can no longer consider "comfort" and "efficiency" In these extreme scenarios, the test indicator can be directly used as a test indicator whether the smart car can avoid the collision in time, and if it is unavoidable, how much the collision speed is controlled. In fact, it is to minimize the damage.

It should be noted that if the driver simulation model is constructed, the key indicator data can also be used for iterative testing of the driver simulation model to optimize the driver simulation model. At this time, the above-mentioned first type of vehicle is the driver simulation model. controlled vehicle. In addition, when constructing the driver simulation model, the first type of vehicle may be a vehicle controlled by a specified control algorithm.

It should also be noted that, in the process of optimizing the driver simulation model, the driver simulation model of each lane changing habit will appear with a certain probability, and a traffic environment model is also formed at this time. After the human driver's continuous iterative cycle test in the whole system, the driver simulation model is getting closer and closer to the driving behavior of the human driver. In turn, the human driver will be in a more and more realistic and complex traffic environment. in the model. Thus, the learning evolution of the driver model and the traffic environment model is completed.

In some other embodiments, in order to test and evaluate the performance of the automatic driving algorithm in the real traffic environment, on the basis of the driving behavior analysis shown in FIG. 1 , the driving behavior analysis method further includes the following steps, and the method flowchart is shown in FIG. 3 . Show:

S50, constructing a traffic environment simulation model based on the driver simulation model.

In this embodiment, the simulation model of the traffic environment can be obtained by incorporating the driver simulation model into the simulated traffic scene, which lays a foundation for subsequent simulation tests.

S60, using the traffic environment simulation model to test the anthropomorphic automatic driving algorithm.

In the process of executing step S60, the autonomous driving vehicle algorithm can be integrated into the traffic environment simulation model. Under safe and efficient conditions, it partially replaces the real vehicle test on public roads.

The driving behavior analysis method provided by the embodiment of the present invention can construct a driver simulation model based on the driving behavior data of the driver for the simulated traffic scene, and then use the driver simulation model to perform anthropomorphic processing on the automatic driving algorithm to be analyzed. Based on the method disclosed in the present invention, it is possible to realize the automatic driving algorithm learning and imitating human driving behavior to the greatest extent possible, thereby helping to improve the safety and comprehensive promotion of the automatic driving vehicle.

Based on the driving behavior analysis method provided by the above-mentioned embodiment, an embodiment of the present invention correspondingly provides a driving behavior analysis device. The schematic structural diagram of the device is shown in FIG. 4 , including:

an acquisition module 10, configured to acquire the driving behavior data of the driver for a preset simulated traffic scene;

The first processing module 20 is configured to process the driving behavior data to obtain key indicator data of the simulated traffic scene;

a building module 30, used for building a driver simulation model according to the key indicator data;

The second processing module 40 is used for anthropomorphizing the automatic driving algorithm to be analyzed by using the driver simulation model; wherein,

The obtaining module 10 is specifically configured to: determine the simulated traffic scene designated by the driver based on the preset environment interactive interface; activate the preset cockpit simulation device; and collect the driving behavior data generated by the cockpit device.

Preferably, the building module 30 is specifically used for:

Determine following behavior data based on key indicator data, and process the following behavior data to obtain a following behavior model; and/or determine lane changing behavior data based on key indicator data, and process the lane changing behavior data to obtain a lane changing behavior model.

Preferably, the building module 30 is also used for:

The traffic environment simulation model is constructed based on the driver simulation model.

Preferably, the second processing module 40 is further used for:

The anthropomorphic automatic driving algorithm is tested by using the traffic environment simulation model.

The driving behavior analysis device provided by the embodiment of the present invention can construct a driver simulation model based on the driving behavior data of the driver for the simulated traffic scene, and then use the driver simulation model to perform anthropomorphic processing on the automatic driving algorithm to be analyzed. Based on the device disclosed in the present invention, it is possible to realize the automatic driving algorithm learning and imitating human driving behavior to the greatest extent possible, thereby helping to improve the safety and comprehensive promotion of the automatic driving vehicle.

The driving behavior analysis method and device provided by the present invention have been described in detail above. The principles and implementations of the present invention are described in this paper by using specific examples. The descriptions of the above embodiments are only used to help understand the present invention. method and its core idea; at the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation and application scope. Invention limitations.

It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. For the same and similar parts among the various embodiments, refer to each other Can. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method.

It should also be noted that in this document, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply those entities or operations There is no such actual relationship or order between them. Furthermore, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article, or device of a list of elements is included, inherent to, or is also included for, those processes. , method, article or device inherent elements. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

The above description of the disclosed embodiments enables any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (8)

1. A driving behavior analysis method, characterized by comprising:

acquiring driving behavior data of a driver aiming at a preset simulated traffic scene; wherein, the vehicles in the simulated traffic scene are divided into two types: the first type is a vehicle controlled by a human driver through a cockpit simulation device, and the second type is a vehicle controlled by a non-human driver;

processing the driving behavior data to obtain key index data of the simulated traffic scene;

constructing a driver simulation model according to the key index data;

carrying out personification processing on an automatic driving algorithm to be analyzed by utilizing the driver simulation model; wherein,

the acquiring of the driving behavior data of the driver for the preset simulated traffic scene includes:

determining a simulated traffic scene designated by a driver based on a preset environment interactive interface;

activating a preset cockpit simulation device;

and collecting driving behavior data generated by the cockpit simulation device.

2. The method of claim 1, wherein said building a driver simulation model from said key indicator data comprises:

determining following behavior data based on the key index data, and processing the following behavior data to obtain a following behavior model; and/or

And determining lane changing behavior data based on the key index data, and processing the lane changing behavior data to obtain a lane changing behavior model.

3. The method of claim 1, further comprising:

and constructing a traffic environment simulation model based on the driver simulation model.

4. The method of claim 3, further comprising:

and testing the automatic driving algorithm after anthropomorphic processing by using the traffic environment simulation model.

5. A driving behavior analysis device characterized by comprising:

the acquisition module is used for acquiring driving behavior data of a driver aiming at a preset simulated traffic scene; wherein, the vehicles in the simulated traffic scene are divided into two types: the first type is a vehicle controlled by a human driver through a cockpit simulation device, and the second type is a vehicle controlled by a non-human driver;

the first processing module is used for processing the driving behavior data to obtain key index data of the simulated traffic scene;

the construction module is used for constructing a driver simulation model according to the key index data;

the second processing module is used for carrying out anthropomorphic processing on the automatic driving algorithm to be analyzed by utilizing the driver simulation model; wherein,

the acquisition module is specifically configured to: determining a simulated traffic scene designated by a driver based on a preset environment interactive interface; activating a preset cockpit simulation device; and collecting driving behavior data generated by the cockpit simulation device.

6. The apparatus according to claim 5, wherein the building block is specifically configured to:

determining following behavior data based on the key index data, and processing the following behavior data to obtain a following behavior model; and/or determining lane changing behavior data based on the key index data, and processing the lane changing behavior data to obtain a lane changing behavior model.

7. The apparatus of claim 5, wherein the build module is further configured to:

and constructing a traffic environment simulation model based on the driver simulation model.

8. The apparatus of claim 7, wherein the second processing module is further configured to:

and testing the automatic driving algorithm after anthropomorphic processing by using the traffic environment simulation model.

Images